OK! We're in. So, that wasn't so bad -- just needed to add a bit of code
that set up an eigenvalue problem.
Jmol 11.9.34 does the equivalent of PyMol's "pair_fit" (without the
capability to determine atom-atom matchings a priori or to throw out poor
matches).
The COMPARE command does atom-based or orientation-based
reorientation/superposition. It will be quite interesting to see how this
compares with PyMol, but I think you will find it to be *just* a little
better than PyMol. Maybe. At least VERY VERY CLOSE. (It's possible that
PyMol is automatically removing outliers, I don't know....)
Jmol is using an exact solution that utilizes quaternions, not the older
version that requires iteration that I think Pymol is using. In that sample
I had of yours, Kersten, it's possible that you did not use exactly the atom
set you indicated, but if you did, then Jmol is reporting a very slightly
different match than PyMol, with a slightly smaller RMSD (because it's
exact). References for this are below:
# * see Berthold K. P. Horn,
# * "Closed-form solution of absolute orientation using unit
quaternions" J.
# * Opt. Soc. Amer. A, 1987, Vol. 4, pp. 629-642
# * http://www.opticsinfobase.org/viewmedia.cfm?uri=josaa-4-4-629&seq=0
# *
# * and Lydia E. Kavraki, "Molecular Distance Measures"
# * http://cnx.org/content/m11608/latest/
As part of this work, I've added a new predefined atom set SPINE which
includes only the continuous set of atoms making up the "true" backbone of
proteins and nucleic acid. Specifically, spine includes:
*.CA, *.N, *.C and *.P, *.O3', *.O5', *.C3', *.C4', *.C5'
Note that it does not include *.O1P or *.O2P. (Should it?)
To match your data, I used:
spine or connected(_P)
to get all the atoms you said you used (I think -- please check). But I
think you could leave out those O1P and O2P atoms and be just fine.
So here is what I used. t2.pdb is loaded twice so we can compare.
load files "t1.pdb" "t2.pdb" "t2.pdb"
color {2.1} blue
color {3.1} yellow
frame *
cartoons only
var a={spine or connected(_P)}
compare {2.1} {1.1} ATOMS {77-82 and a} {41-46 and a} {92-100 and a} {27-35
and a} ROTATE TRANSLATE
The keyword ATOMS is optional. Note that, like PyMol, we can indicate
however many pairs of atom sets we want. Jmol will do the proper ANDing with
the models, so you could use for a protein, for example, just:
compare {2.1} {1.1} atoms {*.CA} rotate translate
And it will do what you expect it to do.
If you leave out ROTATE or TRANSLATE, Jmol will report the RMSD before and
after the calculation, but it won't actually do the rotation or translation.
So, for example, for your model there, it reports:
RMSD 0.98 --> 0.95 Angstroms
Apparently Jmol thinks it did better than Pymol. You be the judge.
http://chemapps.stolaf.edu/jmol/docs/examples-11/align/U4_snRNA_jmol_align.jpg
(orange: target; yellow:PyMol; blue: Jmol
another view (Jmol orange; PyMol red, I think):
http://chemapps.stolaf.edu/jmol/docs/examples-11/align/jmol_pymol_compare.jpg
So that's pretty darned close.
The data for that are in:
http://chemapps.stolaf.edu/jmol/docs/examples-11/align/jmol_pymol_compare.pdb
Files can be downloaded at
http://chemapps.stolaf.edu/jmol/docs/examples-11/Jmol-11.zip
Bob
On Tue, Mar 16, 2010 at 7:43 AM, Robert Hanson <[email protected]> wrote:
> Let's go with COMPARE instead of ALIGN. That way we can have all sorts of
> flexibility, anywhere from just reporting orientational or coordinate
> superposition RMSD, doing just the rotation, doing just the translation,
> etc.
>
> Bob
>
>
> On Mon, Mar 15, 2010 at 2:58 PM, Robert Hanson <[email protected]> wrote:
>
>> essentially. Any situation where you can describe orientations.
>> Carbohydrates, for example, could be quite amenable.
>>
>> On Mon, Mar 15, 2010 at 2:42 PM, <[email protected]> wrote:
>>
>>> On Mon, 15 Mar 2010 14:33:08 -0500, Robert Hanson <[email protected]>
>>> wrote:
>>> > "Spacial" doesn't really describe it.
>>> >
>>> > The method goes like this:
>>> >
>>> > --Identify pairs of matched residues (as opposed to atoms).
>>> >
>>>
>>> So not of general use but limited to the special cases of nucleic acids
>>> and proteins.
>>>
>>> Rich
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> Download Intel® Parallel Studio Eval
>>> Try the new software tools for yourself. Speed compiling, find bugs
>>> proactively, and fine-tune applications for parallel performance.
>>> See why Intel Parallel Studio got high marks during beta.
>>> http://p.sf.net/sfu/intel-sw-dev
>>> _______________________________________________
>>> Jmol-users mailing list
>>> [email protected]
>>> https://lists.sourceforge.net/lists/listinfo/jmol-users
>>>
>>
>>
>>
>> --
>> Robert M. Hanson
>> Professor of Chemistry
>> St. Olaf College
>> 1520 St. Olaf Ave.
>> Northfield, MN 55057
>> http://www.stolaf.edu/people/hansonr
>> phone: 507-786-3107
>>
>>
>> If nature does not answer first what we want,
>> it is better to take what answer we get.
>>
>> -- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
>>
>
>
>
> --
> Robert M. Hanson
> Professor of Chemistry
> St. Olaf College
> 1520 St. Olaf Ave.
> Northfield, MN 55057
> http://www.stolaf.edu/people/hansonr
> phone: 507-786-3107
>
>
> If nature does not answer first what we want,
> it is better to take what answer we get.
>
> -- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
>
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
1520 St. Olaf Ave.
Northfield, MN 55057
http://www.stolaf.edu/people/hansonr
phone: 507-786-3107
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Jmol-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jmol-users